← Back to Home

Five-year Leqembi follow-up shows signs of slower decline, but long-term benefits still need more rigorous comparison

Among 351 patients with early Alzheimer’s disease who continued treatment, disease progression after five years was lower than in an external follow-up cohort. However, the lack of a concurrent randomized placebo group limits the interpretation of these long-term data.

By SURL BioNews

For people with Alzheimer’s disease and their families, whether treatment can preserve memory and daily functioning for a few more years matters more directly than short-term differences on assessment scales. Eisai and Biogen presented five-year follow-up results for Leqembi (lecanemab) at the 2026 Clinical Trials on Alzheimer’s Disease (CTAD) conference in Boston, adding data to this question: patients who continued treatment declined less than an external natural-history cohort, but this comparison no longer carries the strength of evidence of the original randomized trial.

Lecanemab is a monoclonal antibody targeting amyloid-beta aggregates. These data come from the Phase 3 Clarity AD trial and its long-term extension; the original trial enrolled 1,795 patients with early disease in an 18-month randomized, double-blind, placebo-controlled study, while the latest analysis covers 351 people who continued treatment for five years. Development partner BioArctic also released the same results.

As assessed by the Clinical Dementia Rating–Sum of Boxes (CDR-SB), 61.9% of the treatment cohort progressed to the next disease stage within five years, compared with 77.5% of the external Alzheimer’s Disease Neuroimaging Initiative (ADNI) cohort; the proportions progressing to moderate or severe dementia were 31.0% and 48.7%, respectively. These figures suggest slower decline, while also showing that continued treatment did not prevent disease progression in all patients.

Results in patients at an earlier stage drew particular attention. Among participants who volunteered for a tau positron emission tomography substudy and had lower tau accumulation in the brain, 66% had no worsening in their CDR-SB scores after five years, of whom 62% improved from baseline. These are assessment-scale results from a specific subgroup and cannot be directly generalized to all patients or equated with reversal of the disease; the announcement did not report the number of participants in this subgroup in the five-year analysis.

The choice of comparator is central to interpreting long-term benefits. The extension study had no concurrent randomized placebo group and had to draw on data from ADNI and another follow-up study, BioFINDER. BioArctic specifically noted that all BioFINDER participants were at the mild cognitive impairment stage, with no participants with mild dementia, and that their baseline CDR-SB scores were also lower. These differences at the outset, along with potential selection bias among those able to continue participating in treatment for five years, limit causal inference.

On safety, both announcements stated that no new safety signals were identified. Among amyloid-related imaging abnormalities, ARIA-E, which involves edema or effusion, occurred mostly during the first six months of treatment, with lower incidence in subsequent years. However, “no new signals” does not mean that existing risks have disappeared; observations in a cohort that continued long-term treatment also cannot fully represent everyone who started the drug.

This study extends the observation period for lecanemab to five years, supporting further evaluation of whether continued treatment can preserve cognition and daily functioning. The magnitude of the benefit and which patients are most likely to benefit still need to be assessed alongside the full analytical methods, the handling of withdrawals and missing data, and more comparable long-term evidence. The conference and company data released so far are insufficient to answer these questions on their own.

References

  1. Eisai and Biogen
  2. BioArctic